本人最近下载了DOTA1.5版本的数据集，在此将处理的流程在此与大家做一个分享：
目的：将DOTA数据集处理为YOLO模型可以进行训练的形式
在拿到手的DOTA数据集标注样式如下：
其各个部分的含义为：
imagesource: GoogleEarth：这表示图像的来源是Google Earth
gsd: 0.146343590398：GSD（Ground Sample Distance）表示地面采样距离，即图像中每个像素所代表的实际地面面积的大小。这里的0.146343590398米/像素意味着图像中的一个像素对应地面上大约0.146米的距离。
937 913 921 912 923 874 940 875 small-vehicle 0：前八个数代表从目标框左上角开始的(x,y）坐标，沿顺时针方向的四点坐标，small-vehicle为类别，后面紧跟的一个数字代表检测的难易程度。
我们首先去除每个标签文件用不上的前两行，代码如下：
import os
def remove_first_two_lines(file_path):
with open(file_path, 'r', encoding='utf-8') as file:
lines = file.readlines()
with open(file_path, 'w', encoding='utf-8') as file:
file.writelines(lines[2:])
directory = 'E:/UserTable/DOTA1.5/val/labelTxt-v1.5' # 指定要处理的文件夹路径，这里使用当前文件夹
for filename in os.listdir(directory):
if filename.endswith('.txt'):
file_path = os.path.join(directory, filename)
remove_first_two_lines(file_path)去除了前两行代码，我们可以使用一下代码对图片和对应标签进行分割：
import cv2
import os
# 图像宽不足裁剪宽度,填充至裁剪宽度
def fill_right(img, size_w):
size = img.shape
# 填充值为数据集均值
img_fill_right = cv2.copyMakeBorder(img, 0, 0, 0, size_w - size[1],
cv2.BORDER_CONSTANT, value = (107, 113, 115))
return img_fill_right
# 图像高不足裁剪高度,填充至裁剪高度
def fill_bottom(img, size_h):
size = img.shape
img_fill_bottom = cv2.copyMakeBorder(img, 0, size_h - size[0], 0, 0,
cv2.BORDER_CONSTANT, value = (107, 113, 115))
return img_fill_bottom
# 图像宽高不足裁剪宽高度,填充至裁剪宽高度
def fill_right_bottom(img, size_w, size_h):
size = img.shape
img_fill_right_bottom = cv2.copyMakeBorder(img, 0, size_h - size[0], 0, size_w - size[1],
cv2.BORDER_CONSTANT, value = (107, 113, 115))
return img_fill_right_bottom
# 图像切割
# img_floder 图像文件夹
# out_img_floder 图像切割输出文件夹
# size_w 切割图像宽
# size_h 切割图像高
# step 切割步长
def image_split(img_floder, out_img_floder, size_w = 1000, size_h = 1000, step = 800):
img_list = os.listdir(img_floder)
count = 0
for img_name in img_list:
number = 0
# 去除.png后缀
name = img_name[:-4]
img = cv2.imread(img_floder + "\\" + img_name)
size = img.shape
# 若图像宽高大于切割宽高
if size[0] >= size_h and size[1] >= size_w:
count = count + 1
for h in range(0, size[0] - 1, step):
start_h = h
for w in range(0, size[1] - 1, step):
start_w = w
end_h = start_h + size_h
if end_h > size[0]:
start_h = size[0] - size_h
end_h = start_h + size_h
end_w = start_w + size_w
if end_w > size[1]:
start_w = size[1] - size_w
end_w = start_w + size_w
cropped = img[start_h : end_h, start_w : end_w]
# 用起始坐标来命名切割得到的图像，为的是方便后续标签数据抓取
name_img = name + '_'+ str(start_h) +'_' + str(start_w)
cv2.imwrite('{}/{}.png'.format(out_img_floder, name_img), cropped)
number = number + 1
# 若图像高大于切割高,但宽小于切割宽
elif size[0] >= size_h and size[1] < size_w:
print('图片{}需要在右面补齐'.format(name))
count = count + 1
img0 = fill_right(img, size_w)
for h in range(0, size[0] - 1, step):
start_h = h
start_w = 0
end_h = start_h + size_h
if end_h > size[0]:
start_h = size[0] - size_h
end_h = start_h + size_h
end_w = start_w + size_w
cropped = img0[start_h : end_h, start_w : end_w]
name_img = name + '_' + str(start_h) + '_' + str(start_w)
cv2.imwrite('{}/{}.png'.format(out_img_floder, name_img), cropped)
number = number + 1
# 若图像宽大于切割宽,但高小于切割高
elif size[0] < size_h and size[1] >= size_w:
count = count + 1
print('图片{}需要在下面补齐'.format(name))
img0 = fill_bottom(img, size_h)
for w in range(0, size[1] - 1, step):
start_h = 0
start_w = w
end_w = start_w + size_w
if end_w > size[1]:
start_w = size[1] - size_w
end_w = start_w + size_w
end_h = start_h + size_h
cropped = img0[start_h : end_h, start_w : end_w]
name_img = name + '_'+ str(start_h) +'_' + str(start_w)
cv2.imwrite('{}/{}.png'.format(out_img_floder, name_img), cropped)
number = number + 1
# 若图像宽高小于切割宽高
elif size[0] < size_h and size[1] < size_w:
count = count + 1
print('图片{}需要在下面和右面补齐'.format(name))
img0 = fill_right_bottom(img, size_w, size_h)
cropped = img0[0 : size_h, 0 : size_w]
name_img = name + '_'+ '0' +'_' + '0'
cv2.imwrite('{}/{}.png'.format(out_img_floder, name_img), cropped)
number = number + 1
print('{}.png切割成{}张.'.format(name,number))
print('共完成{}张图片'.format(count))
# txt切割
# out_img_floder 图像切割输出文件夹
# txt_floder txt文件夹
# out_txt_floder txt切割输出文件夹
# size_w 切割图像宽
# size_h 切割图像高
def txt_split(out_img_floder, txt_floder, out_txt_floder, size_h = 1000, size_w = 1000):
img_list = os.listdir(out_img_floder)
for img_name in img_list:
# 去除.png后缀
name = img_name[:-4]
# 得到原图像(也即txt)索引 + 切割高 + 切割宽
name_list = name.split('_')
txt_name = name_list[0]
h = int(name_list[1])
w = int(name_list[2])
txtpath = txt_floder + "\\" + txt_name + '.txt'
out_txt_path = out_txt_floder + "\\" + name + '.txt'
f = open(out_txt_path, 'a')
# 打开txt文件
i=0
with open(txtpath, 'r') as f_in:
lines = f_in.readlines()
# 逐行读取
for line in lines:
if i<2:
i = i+1
else:
splitline = line.split(' ')
label = splitline[8]
difficult = splitline[9]
x1 = int(float(splitline[0]))
y1 = int(float(splitline[1]))
x2 = int(float(splitline[2]))
y2 = int(float(splitline[3]))
x3 = int(float(splitline[4]))
y3 = int(float(splitline[5]))
x4 = int(float(splitline[6]))
y4 = int(float(splitline[7]))
if w <= x1 <= w + size_w and w <= x2 <= w + size_w and \
w <= x3 <= w + size_w and w <= x4 <= w + size_w and \
h <= y1 <= h + size_h and h <= y2 <= h + size_h and \
h <= y3 <= h + size_h and h <= y4 <= h + size_h:
f.write('{} {} {} {} {} {} {} {} {} {}'.format(int(x1 - w),
int(y1 - h), int(x2 - w), int(y2 - h), int(x3 - w),
int(y3 - h), int(x4 - w), int(y4 - h),
label, difficult))
f.close()
print('{}.txt切割完成.'.format(name))
# 图像数据集文件夹
img_floder = r'E:/UserTable/DOTA1.5/val/images/images'
# 切割得到的图像数据集存放文件夹
out_img_floder = r'E:/UserTable/DOTA1.5/val/images/images_qiege'
# txt数据集文件夹
txt_floder = r'E:/UserTable/DOTA1.5/val/labelTxt-v1.5'
# 切割后数据集的标签文件存放文件夹
out_txt_floder = r'E:/UserTable/DOTA1.5/val/labels_qiege'
# 切割图像宽
size_w = 1024
# 切割图像高
size_h = 1024
# 切割步长,重叠度为size_w - step
step = 824
image_split(img_floder, out_img_floder, size_w, size_h, step)
txt_split(out_img_floder, txt_floder, out_txt_floder, size_h, size_w)
最后，使用以下代码完成YOLO标注形式的数据集标签转换：
import os
from PIL import Image
def convert_dota_to_yolo(dota_folder):
yolo_folder = "E:/UserTable/DOTA1.5/val/labels_yolo" # 存储转换后的yolo格式文件夹
made = "E:/UserTable/DOTA1.5/val/images/images_qiege"
os.makedirs(yolo_folder, exist_ok=True)
label_map = {'airport':0,
'small-vehicle':1,
'large-vehicle':2,
'plane':3,
'storage-tank':4,
'ship':5,
'harbor':6,
'ground-track-field':7,
'soccer-ball-field':8,
'tennis-court':9,
'swimming-pool':10,
'baseball-diamond':11,
'roundabout':12,
'basketball-court':13,
'bridge':14,
'helicopter':15,
'container-crane':16,
'helipad':17}
label_index = 1
for filename in os.listdir(dota_folder):
if filename.endswith(".txt"):
dota_file = os.path.join(dota_folder, filename)
yolo_file = os.path.join(yolo_folder, filename.replace(".txt", ".txt"))
caonimade = os.path.join(made, filename.replace(".txt", ".png"))
with open(dota_file, "r") as dota_f:
with open(yolo_file, "w") as yolo_f:
for line in dota_f:
line = line.strip().split()
class_name = line[8]
x1, y1, x2, y2, x3, y3, x4, y4 = map(float, line[:8])
# 如果类别名称不在字典中，则添加新的索引
if class_name not in label_map:
raise ValueError("有未命名的东西,giao!!!")
width, height = get_image_size(caonimade)
# 计算yolo格式的中心点坐标和宽高
x_center = (x1 + x3) / 2 / width
y_center = (y1 + y3) / 2 / height
width = (x3 - x1) / width
height = (y3 - y1) / height
# 将转换后的结果写入yolo格式文件
yolo_f.write(f"{label_map[class_name]} {x_center} {y_center} {width} {height}\n")
print("转换完成！")
def get_image_size(image_path):
img = Image.open(image_path)
width, height = img.size
return width, height
# 调用函数进行转换
convert_dota_to_yolo("E:/UserTable/DOTA1.5/val/labels_qiege")
就此完成处理的全部流程！
完整代码同样打包在了：https://download.csdn.net/download/wbk1669177085/89888723，可以免费下载
扫一扫
1809
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
